Southern Ocean - Results and Discussion .1 Pacific Ocean

2.3 Results and Discussion .1 Pacific Ocean

2.3.3 Southern Ocean

A further complication of the relationship between xs231Pa1xs230Th ratio and particle flux is apparent in results from the Southern Ocean. The first investigation of ^Pa and 23OTh in this region was carried out by DeMaster (1979). In rapidly accumulating (up to 100cm/ka) siliceous sediments within the Antarctic Circumpolar Current (ACC) between 48OS and 55OS, DeMaster found evidence for preferential deposition of ^Pa relative to 230Th, with inventories of ^OTh and ^Pa exceeding the production rate in the water column by up to 6 and 14 times, respectively.

Moreover, he showed that the 23lPal23oTh deposition ratio was up to five times higher than the production ratio of 0.093. DeMaster explained his findings by a boundary scavenging effect and hypothesized that, due to the high particle flux within the ACC, scavenging of ^Pa and ^Th is so effective that both

radionuclides are quantitatively stripped from the water column. This would imply that the high xs231Pa/xs230Th ratios in sediments underlying the ACC reflect both the local production of the two radionuclides in the water column and the remaining fractions of dissolved 231Pa and (less important) of 2307% imported from the Atlantic Ocean by advection (Yu et al., 1996).

Several recent findings indicate that a simple boundary scavenging model cannot be directly applied in the Southern Ocean. First, total mass fluxes measured in sediment traps located at 50Â° (PF) and 55OS (B0112) within the ACC are 38.3g-m-

2.y-1 (Wefer and Fischer, 1991) and 53g-m-2.y-1 (Fischer and Wefer, unpublished), respectively. Although not corrected for trapping efficiency, these are much lower fluxes than measured in upwelling regions such as the Panama Basin or the California margin ( > I 0Og.m-2.y-1; Honjo, 1982; Yu, 1994). These low fluxes are also in agreement with primary production being much lower (in general less than lOOgC-m-2-y-I), than in upwelling regions (sometimes exceeding 200gC.m-2Â¥y-1 Fig.2-4 a,b; Fig.2-5 a,b). According to Fig.2-3, such mass flux would be insufficient to enhance 231Pa scavenging above its rate of production in the water column.

Second, according to the reversible scavenging model of Bacon and Anderson (1982), the concentrations of ^Pa and ^Th in the water column should adjust to the flux of particulate matter. Consequently, the high particle fluxes postulated by DeMaster (1979) should lead to a depletion of their dissolved activities relative to regions with low particle fluxes. Actually, there are no indications of a depletion of 231Pa and ^Th in the water column within the ACC relative to the less productive regions to the north and south (Rutgers van der Loeff and Berger, 1993; Walter et al., 1997). Third, a recent study of surface sediments in the Atlantic sector of the Southern Ocean (Fig.2-5) has revealed that the region of high xs^Palxs230Th ratios is not restricted to the area of the ACC, but persists throughout the Weddell Sea (Walter et al., 1997). This is in contrast to the low particle fluxes of this region compared to the ACC, inferred from sediment trap data (< 5Og-m-2-y-l; Fischer et al.

1988; Wefer and Fischer 1991) and fluxes of biogenic opal through the sediment- water interface deduced from pore water measurements of silicate (SchlÃ¼te et al., submitted). The latter observation unequivocally illustrates the decoupling between xs231P&230Th ratio and the mass flux of particulate matter in this region. As boundary scavenging cannot explain the observations above, another factor than the mass flux must be responsible for the high xs^Pa/xs230Th ratios in the sediments within and south of the ACC.

As discussed above, the application of the xs231Palxs2^Th ratio as a paleoproductivity proxy requires a constant 23OThl231Pa fractionation factor (F) with a strong preference for the adsorption of 230Th over 23lPa on particles. The F-factor is defined as

The few data on F-factors available from the literature (Anderson et al., 1983a;

1983b; Nozaki and Nakanishi, 1985; Scholten et al., 1995) show that this assumption holds for large areas of the world's oceans. However, in the Atlantic sector of the Southern Ocean it has been shown to be no longer valid (Rutgers van

der Loeff and Berger, 1993; Walter et al., 1997). Typical Open ocean values of F around 10 (Anderson et al., l983a; 1983b), indicating a strong preference for the adsorption of 23OTh relative to 23lPa on particles, were only observed north of 48's (Fig.2-6). To the south, F decreases strongly and implies a change in the scavenging preference of 230Th over 231Pa, either due to a southward decrease in the scavenging efficiency of ^Th, an increase in the scavenging efficiency of 231Pa, or both.

The N-S decrease in F (Rutgers van der Loeff and Berger, 1993; Walter et al,, 1997) is well correlated with a respective increase in the content of biogenic opal on sedimenting particles (Fischer et al., 1988; Wefer and Fischer, 1991). An experimental study on the adsorption of ^Th and 23lPa on different solid phases by Anderson et al. (1992) has shown that amorphous silica does not fractionate 23lPa and 230Th (F = 1.1; Fig.2-7), possibly because of its high affinity for 231Pa (DeMaster, 1979; Taguchi et al., 1989; Anderson et a l , 1992; Lao et al., 1992b;

Kumar et al., 1995). The increased predominance of opal may explain the latitudinal decrease in F. Support for this suggestion Comes from the analysis of xs23-IPa/230Th ratios in settling particles (collected by sediment traps), where estimated F values are negatively correlated with the content of opal of these particles (Tab.2-4).

0 4 . , . i . * . ~ . , . . . \

-40 45 50 55 60 65 70 75

latitude south

Fig.2-6: Plot of the 230ThPPa fractionation factor F against latitude in the Atlantic sector of the Southern Ocean (from Walter et al., 1997).

Based On these observations it is concluded that biogenic opal might explain the enhanced scavenging of 23lPa relative to ^Th south of the Polar Front, at least in the Atlantic sector of the Southern Ocean (Walter et al., 1997). This would imply that in regions where the sedimenting flux is dominated by biogenic opal, the xs231Pa/xs230Th ratio is no more a reliable indicator for the mass flux of particles.

Table 2-4: Opal contents and xs23~Palxs230Th ratios of sinking particles, and estimated F- factor (from Walter et al., 1997)

trap latitude longitude depth mass flux opal xs231p~ *(231p&30~h) F-factor (m) a.m-2.a-1 (%) 2 3 b h dissolved

PF1-450Â°090' 05'464'E 700 3 8 3 40 0 3 2 2 i 0 0 6 6 0 5 9 9 i 0 0 6 4 1 8 6 5 0 4 3

B01+254Â°202' 03Â°202' 450 5 3 1 62 0 3 4 4 k 0 0 3 4 0 4 6 5 i 0 0 3 6 1 3 5 5 0 1 7 WS 3 64" 54 I ' S 02'33 8 ' W 360 33 7 70 0 2 7 8 k 0 016 0 280 i 0 058 I 01 k 0 23

Data On mass flux and opal content of PF 1-4 and WS-3 are from Wefer and Fischer (1991), and of B 0 1+2 from Wefer and Fischer ~un~ublished~.

Errors are IG propagated from counting staiistics andblank.

*interpolated from latitudinal gradient of dissolved 231 PaPOTh ratios

" .

Mn02 Sed. Sed. Fe203 Fe203 A1203 Silica

Fig.2-7: 2aThP31Pa fractionation factors for different particle types, determined in ariificial seawater, as weil as for Fe203 and natural sediments in filtered seawater (FSW). Reproduced with permission from Anderson et al. ( I 992).

In the Southern Ocean, where high flux rates of biogenic opal to the sediment also have occurred in the past, as is weil documented by the circumantarctic opal belt (DeMaster, 1981), it is suggested that the use of the xs231Pa/xs230Th ratio as a paleoproductivity proxy is strongly limited (Walter et al., 1997). Variations of the xs231Padxs230Tho ratio through time, usually interpreted as reflecting changes in the total mass flux of particles, could also result from changes in the content of biogenic opal On sinking particles.

2.3.4 Weddell

Over most of the oceans the depositional flux of 23OTh approximately equals its local rate of production with little net lateral transport (Anderson et al., 1983a; 1983b;

Bacon et al., 1985; Suman and Bacon, 1989; Francois et al., 1990; Yu, 1994). Even in the Atlantic Ocean north of 50Â°S despite the strong effect of advection, 85-90Y0 of the 23OTh produced is still scavenged within the Atlantic and only 10-1 are advected to the south (Yu et al., 1996). lf the flux/production ratio for 230Th remains still close to I in the Southern Ocean (or slightly above I due to advection of dissolved 23OTh from the Atlantic), the mass balance for 231Pa can be calculated (Yu et al,, 1996). By taking a mean xs231Pa/xs230Th rain ratio of 0.165 for the region south of 50Â°S they estimated that the depositional flux of 23lPa throughout the entire Southern Ocean balances in situ production plus 237Pa import from the Atlantic Ocean, suggesting this region to be an important sink for 23lPa.

The assumption of a constant flux/production ratio of 23OTh around I does not appear to be valid throughout the entire Southern Qcean. Based On measurements of the dissolved and particulate concentrations of 23OTh in the water column of the South Atlantic, Rutgers van der Loeff and Berger (1993) hypothesized that in the Central Weddell Sea south of the Antarctic Weddell boundary (AWB), scavenging rate of 23OTh might be strongly reduced relative to the regions $0 the north (ACC).

They based their suggestion On both the extremely low particle fluxes (Fischer et al., 1988; Schlueter et al., submitted) and the short residence time of water in the Weddell Sea (35 years; Rutgers van der Loeff and Berger, 1993) comparable to the scavenging residence time of 230Th. Further support for low scavenging rates in the Central Weddell Sea comes from the sedimentary record, with extremely low 210Pb inventories of only 15Y0 of its production rate (Rutgers van der Loeff and Berger, 1991). These findings are confirmed by a still ongoing investigation of the inventories of 237Pa and 23OTh, in three sediment cores from the Western Central Weddell Sea (Walter et al., in prep.). The selected area is characterized by slow bottom currents, so that effects of lateral sediment redistribution are only small, which is a reflection of sluggish circulation of the Center of the Weddell Gyre compared to the outer Parts such as the northern Weddell (Pudsey et al., 1988).

Therefore, it is not to be expected that the core locations are influenced by sediment winnowing (Pudsey et al., 198%). Results from the core P S I 5 0 8 (67'00. l'S/32Â°23.5'W have shown that over the last 150ka only 37% and 52Y0 of the production of 23OTh and 237Pa, respectively, are found in the sediment. These observations imply a N-S Change in the scavenging efficiency not only for 231 Pa, but also for 23OTh. lf this observation holds for the entire Weddell Sea, this would suggest that most of the 23OTh and 23Pa produced in this basin must be advected to other regions of the Southern Ocean prior to scavenging, such as the ACC where particle fluxes are higher. Hence, the Weddell Sea would not be a sink for 231Pa, and the high average xs231Pa/xs230Th ratios of 0.15 found in surface sediments (Fig.2-5 a) reflect an increase in the scavenging efficiency of 231Pa relative to 230Th, probably related to the high opal contents of sedimenting particles.

Further evidence for an advective export of 23OTh of up to 50% of the production comes from other ocean basins with short residence times of deep waters ( < I 0 0 years), such as the North Atlantic and the Arctic Seas (Moran et al., 1995; Moran et

al., 1997; Vogler et al., submitted). The export of 23OTh has to be taken into account when using its flux to the sediment as a reference to predict depositional fluxes of 23lPa (e.g. Anderson et al., 1994; Yu, 1996) in such regions.

nhanced scavenging of 23lPa by Mn- and Fe-oxides

In our simple box model illustration, we have shown that the large-scale distributions of 23lPa and 23OTh in oceanic surface sediments can be explained by assuming a homogenous particle composition with a strong preference for the adsorption of 23OTh relative to 233Pa. This assumption is reasonable, provided that scavenging of both radionuclides is primarily determined by the mass flux of particles. There are exceptions, however, as shown for the Southern Ocean. There is experimental evidence that as for opal, fractionation between 23lPa and 23OTh is also limited On manganese oxides and hematite, with 23OThI23lPa fractionation factors of 0.8 and 2.1, respectively (Fig.2-7). Hydrothermal plumes emanating in the vicinity of mid-ocean ridges are an important localized source for Mn- and Fe- oxides. They are effective scavengers for 23lPa and, to a lesser extent, for 23OTh relative to the surrounding area (German et al., 1991; Frank et al., 1994). This so- called hydrothermal scavenging leads to high xs23~Pa/xs230Th ratios in ridge sediments, as found in the vicinity of the East Pacific Rise, (Shimmield and Price, 1988, See also Fig.2-4 a,b). Further evidence for enhanced scavenging of 23Pa relative to 23OTh is reported from the Panama and Guatemala basins and has been attributed to the presence of Mn02 coated particles, originating from recycling of reduced Mn from suboxic sediments (Anderson et al., 1983b). The use of the xs231Pa/xs230Th ratio as a paleoproductivity index is thus further limited near mid- ocean ridges and possibly in regions with diagenetic remobilization of Mn and near mid-ocean ridges,

As we have shown in the discussion above, the xs231Palxs230Th ratio of oceanic surface sediments is determined by three different parameters: the mass flux of particles, their chemical composition and the hydrography. Consequently, it can only be applied as a reliable tracer for bioproductivity in the past, if the influence of the chemical composition and the hydrography is small. This condition appears to be met in the Pacific Ocean, where the xs231Pa/xs230Th ratio is weil correlated with paflicle flux and primary production. This region is an appropriate study area to apply the x s ~ ~ ~ P a o / x s 2 3 ~ T h o ratio as an index for oceanic productivity in the past.

Even under such circumstances, however, the xs231Pa/xs230Th ratio can only provide relative intensities of export flux between low and high productivity regions.

Throughout the modern Atlantic Ocean, there is still a relationship between xs231Palxs230Th ratio and mass flux of particles, but the overall values are lower due to the low residence time of deep water in this ocean basin. Nevertheless, during time periods with similar hydrographic conditions, the xs23~Padxs230Tho ratio should reflect relative changes in paleoproductivity. Conversely, mean basin-wide xs231Padxs230Tho ratios can be used as a paleocirculation tracer (Yu et al., 1996).

Periods of reduced global thesmohaline circulation would be characterized by higher mean x s ~ ~ ~ P a o / x s ~ ~ ~ T h o raiios in Atlantic sediments and higher contrasts in the xs231Pao/xs230Tho ratios between open ocean and ocean margin (as documented in today's Pacific).

In the Southern Ocean these is no more a simple relationship between

xs231 Palxs230Th ratio and mass flux of particulate matter, This is a result of the strong latitudinal decrease in the scavenging preference of 23OTh over 231Pa, probably related to the high contents of biogenic opal of sinking particles. As a consequence, in this region the xs231Pa/xs230Th ratio cannot be applied as a reliable paleoproductivity proxy, as variations of the x s ~ ~ ~ P a o / x s ~ ~ ~ T h o ratio through time, usually interpreted to reflect changes in the total mass flux of particles, could also be explained by changes in the content of biogenic opal On sinking particles. Th@

high affinity of 23lPa to opal, however, could possibly be used to trace fluxes of biogenic opal to the sediment !hat have been totally dissolved (e.g. throughout the Weddell Sea, Walter et al., 1997). Nearly unpreferential scavenging of 230Th and 231Pa by manganese- and iron- oxides further limits the use of the xs231Pa/xs230Th ratio neas mid ocean ridges and in areas of diagenetic remobilization of manganese.

In conclusion, the xs231Pa/xs230Th ratio can serve as a proxy for relative changes in paleoproductivity if, in the time interval of interest, changes in the basin ventilation rate and differential scavenging of both radionuclides due to changes in the chemical composition of pariiculate matter are not significant. The interpretation of the xs231Pa/xs230Th ratio thus requires synoptic information On its basin-wide distribution and On past changes in composition of sedimenting particles. Reliable reconstructions of changes in paleoproductivity, however, can only be made in combination with other independent proxies, (like biogenic barium, authigenic uranium).

Recent improvements in mass spectrometric techniques (e.g. Chen et al., 1992;

Cheng et al., 1996) now allow the study of the distributions of 23OTh and 231Pa in the ocean in more detail. Small sample sizes of only 1-2 liters of seawater (compared to cubic meters with the traditionally alpha spectrometry), sufficient for the high precision determination of 23OTh (Moran et al., 1995; Moran et aLJ 1997; Voglsr et al., submitted) and 23lPa (Francois, Pers. comm,; Vogler Pers. comm.; Edmonds et al., unpubl. data), promise the detailed investigation of scavenging of both radionuclides even in the surface ocean, where their concentrations are extremely low. This new analytical technique could thus provide the required answers to some of the remaining open questions regarding the oceanographic behaviour of 23OTh and 231 Pa.

Table 2-2: Compiled data of ~ $ 3 1 P a l X s 2 3 O T h ratios of Holocene Pacific sediments.

Table 2-2: continued

Table 2-3: Compiled data of ~ $ 3 1 PaIxs23OTh ratios of Holocene Atlantic sediments.

Table 2-3: continued

nhanced scavenging of ^Pa relative to 230Th in the South Atlantic south of the Polar Front: Implications for the use of the 23lPal230Ti-1 ratio as a paleoproductivity proxy

H.J. Walter, M.M. Rutgers van der Loeff and H. HÃ¶ltze

Alfred Wegener Institute for Polar and Marine Research, Bremerhaven, Germany

3.1 Abstract

The fractionation of ^Th and ^Pa was investigated throughout the Atlantic sector of the Southern Ocean. Published scavenging models generally assume that the 231PaI230Th ratio of surface sediments is primarily determined by the mass flux of particles. This relationship holds north of the Polar Front, where low primary productivity coincides with ratios of unsupported 231Pa/230Th xs(231Pa1230Th) in surface sediments below the production ratio of both radionuclides in the water column. However, we observed high xs231Pa/xs230Th ratios, conventionally interpreted as a high-productivity signal, in surface sediments south of Polar Front, especially throughout the Weddell Sea, in contradiction with the low particle flux of this region. Measurements of both dissolved and particulate fractions of ^Pa and

^Th in the water column revealed a strong N-S decrease of the ThIPa- fractionation factor from typical Open ocean values around 10 north of the Polar Front to values between 1 and 2 south of 60's. This observation clearly indicates that the high xs231Pa/xs230Th ratios in surface sediments south of the Antarctic Circumpolar Current are produced by a N-S increase in the relative scavenging efficiency of ^Pa relative to ^Th, most probably due to a change in the chemical composition of particulate matter, and not by a high mass flux. It is speculated that biogenic opal, suggested not to significantly fractionate 231Pa and may explain the enhanced scavenging of ^Pa to the south. This assumption is further supported by extremely high 23lPa1230Th ratios up to 0.34 in material collected with sediment traps south of the Polar Front, where fluxes are primarily determined by biogenic opal, Based on these results we conclude that in regions where the sedimenting flux is dominated by biogenic opal, the 231Pd230Th ratio is not a reliable indicator for the mass flux of particles, thus limiting its use 2s a paleoproductivity proxy in the Southern Ocean.

3.2 Introduction

Oceanic bioproductivity is considered to be an important factor controlling the partitioning of CO2 between deep ocean and atmosphere, which is termed as the 'biological pump" (Eppley and Peterson, 1979; Broecker and Peng, 1982; Martin, 1990). In the euphotic Zone of the surface ocean, CO2 is taken up by phytoplankton and converted into organic matter, which then may sink to the deep ocean. Attempts have been made to explain variations in the atmospheric concentration of CO2 in the past by differences in the bioproductivity of the oceans (Martin, 1990; Sarmiento and Toggweiler, 1984; Keir, 1990). Recently it was argued that iron as limiting factor for today's Southern Ocean productivity, was supplied in glacial periods by wind-

Im Dokument r the use (Seite 27-41)